We study nutrient control of gene expression at the transcriptional and translational levels in the yeast Saccharomyces, focusing on a regulatory mechanism that induces most genes encoding amino acid biosynthetic enzymes in response to starvation for an amino acid. The transcriptional activator in this pathway, GCN4, is induced at the translational level in starved cells on phosphorylation of translation initiation factor 2 (eIF2) by the protein kinase (PK) GCN2. Phosphorylation of eIF2 reduces the concentration of the ternary complex (TC) containing eIF2, GTP, and initiator methionyl tRNA, that transfers tRNAiMet to the 40S ribosome. This impedes general protein synthesis while inducing GCN4 translation. GCN2 is activated in starved cells by binding of uncharged tRNA to a histidyl-tRNA synthetase (HisRS)-like region that functions as a sensor of amino acid limitation. Interactions between the PK and C-terminal domain (Cterm) in GCN2 prevent tRNA binding and kinase activation by basal levels of uncharged tRNA in nonstarved cells. Point mutations in the PK domain were isolated that permit kinase function in the absence of tRNA binding and the positive regulators GCN1 and GCN20. These mutations remove an inhibitory structure in the PK domain that is normally relieved by interaction with the HisRS-N domain bound to tRNA. Phosphorylation of Ser-577 in GCN2 by another kinase decreases the affinity of GCN2 for tRNA and impedes kinase activation in nutrient-replete cells. Dephosphorylation of Ser577, with attendant activation of GCN2, occurs in cells treated with rapamycin, resulting from inhibition of TOR1 and TOR2. The TOR proteins are inactivated in cells starved for nitrogen or carbon, leading to reduced ribosome synthesis and decreased eIF4F activity. Our results show that inhibition of TOR also impedes eIF2 function by activating GCN2 independently of amino acid limitation and elevated levels of uncharged tRNA. The yeast YIH1 protein when overexpressed inhibits GCN2 in amino acid starved cells by competing with the N-terminus of GCN2 for binding to the positive effector GCN1. Deletion of YIH1 does not activate GCN2 in nonstarved cells, however, indicating that YIH1 is not a general negative regulator of GCN2. At native expression levels, YIH1 is in a stoichiometric complex with actin, suggesting that its true function is related to the actin cytoskeleton. Perhaps native YIH1 inhibits GCN2 only under conditions where its association with actin is disrupted. Yeast eIF3 contains 5 core subunits and a sixth more loosely associated protein known as HCR1. From protein interaction assays, we constructed a subunit interaction map for eIF3 and found that eIF3c/NIP1 contains a binding site for eIFs 1 and 5. eIF5, the GTPase activating protein for the TC, and eIF1 both function in selection of AUG as the start codon. The C-terminal segment of eIF5 (CTD), which binds NIP1, also interacts with the beta subunit of eIF2, and these interactions occur simultaneously in vivo, stabilizing a multifactor complex (MFC) containing eIFs 1,2,3, 5 and tRNAiMet. As an eIF5 mutation that disrupts these interactions (tif5-7A) impairs translation in vivo, the MFC appears to be an important translation intermediate. We have incorporated an affinity tag into the 3 largest eIF3 subunits and deleted predicted binding domains in each tagged protein. By purifying and characterizing the mutant subcomplexes, we confirmed all binding interactions predicted by our structural model and uncovered a direct contact between the CTD of eIF3a/TIF32 and beta subunit of eIF2. Overexpressing a CTD-less form of TIF32 exacerbated the initiation defect of the tif5-7A mutation, which weakens the NIP1/eIF5/eIF2 connection. Thus, the two independent eIF2-eIF3 contacts make additive contributions to the efficiency of translation in vivo. Overexpressing the NIP1-NTD sequestered eIF1/eIF5/eIF2 in a defective subcomplex that derepressed GCN4 mRNA translation (Gcd- phenotype) independently of eIF2 phosphorylation. This Gcd- phenotype was suppressed by overproducing the TC, providing the first evidence that association with eIF3 promotes binding of TC to 40S ribosomes in vivo. From genetic and biochemical characterization of mutations in eIF3a/TIF32 and eIF3b/PRT1, we obtained evidence that eIF3 has an additional, rate-limiting function in ribosomal scanning or AUG recognition, following assembly of the 48S initiation complex. By its homology to bacterial IF1, eIF1A is predicted to bind to the decoding (A) site of the 40S ribosome. eIF1A interacts with eIF5B, the ortholog of bacterial IF2, which promotes 40S-60S subunit joining. The fact that overexpressing eIF1A exacerbated the growth defect of a strain lacking eIF5B suggested that its interaction with eIF5B promotes release of eIF1A from the A-site following subunit joining. Consistent with this idea, we mapped the binding domain for eIF5B to the last 24 residues of eIF1A and found that deleting these residues impairs translation in vivo. A larger C-terminal truncation in eIF1A derepresses GCN4 translation in a manner suppressed by overexpressing the TC, providing evidence that eIF1A additionally promotes TC binding to 40S ribosomes in vivo. We have investigated the requirements of GCN4 for transcriptional coactivators by testing mutants from the Saccharomyces Deletion Project for defects in activation by GCN4 in vivo. Our data confirm that GCN4 requires SAGA, SWI/SNF and SRB/mediator, and identify the key nonessential subunits in these complexes required for activation in vivo. Activation by GCN4 also requires the CCR4/NOT complex, RSC and the Paf1 complex, but not ISW1/ISW2. RSC and ISW are ATP-dependent chromatin remodeling complexes related to SWI/SNF, the Paf1 complex is an alternative form of mediator, and CCR4/NOT has been implicated in controlling TBP binding and holoenzyme function. In vitro binding experiments indicate that the GCN4 activation domain binds specifically to CCR4-NOT and RSC in addition to SAGA, SWI/SNF and SRB/MED, but not to the Paf1 complex. Chromatin immunoprecipitation experiments show that GCN4 recruits all six of these coactivators to one of its target genes in living cells. Thus, GCN4 requires a multiplicity of coactivators for activation of individual target genes in vivo.